Presently, there are many network systems having a loop architecture. Such networks preferably are arranged to allow one or more disk drives to be removed from the loop while maintaining operation of the loop. For this reason, disk drives are bypassed with a bypass circuit so that a serial loop may still transfer a signal in the event one of the disk drives is removed. When a disk drive is removed, the associated bypass circuit senses removal of the disk drive and allows data to transfer through the bypass circuit.
Bypassing a disk drive may have a detrimental effect on the performance of the loop. This is a result of the functional differences between the disk drive path and the bypass path. When the disk drive is present, the communications protocol present in the disk drive provides freshly tuned data. Thus, signal integrity of the outbound path is clean. Due to economical considerations, the bypass circuit may not provide this level of function resulting in an outbound signal which is degraded from that of the actual disk drive outbound signal. In the event that plural, serially connected disk drives are inactive and plural bypass circuits are active, signal deterioration is compounded. Some examples of signal deterioration are propagation delay, jitter, and performance degradation, as well as a reduced distance allowed between disk drives. Even if a more costly bypass circuit which includes retiming is used, there still can be unnecessary latency through the communication path.